Labour mobility effect on wages: the professional football players

Paper to be presented at the DRUID Academy 2013
on
DRUID Academy 2013
at Comwell Rebild Bakker, Rebild/Aalborg
Labour mobility effect on wages: the professional football players? case
Antonio Sergio Ribeiro
Technical University of Lisbon, Portugal
Center for Management Studies of Instituto Superior Técnico
[email protected]
Francisco Lima
Technical University of Lisbon
Center for Management Studies of Instituto Superior Técnico
[email protected]
Abstract
In their seminal paper, Lazear and Rosen (1981) consider a setting characterized by tournaments. The high wage
dispersion within football teams can be a sign that clubs set up an internal tournament among players as an incentive
device (Rosen 1981). In the context of football teams, considering that the player?s output is a random variable whose
distribution is controlled by the player himself, club managers may observe the team and player output but cannot
ascertain the extent to which it is due to good fortune, to effort, or to both. Sport industry is a particular example in which
the organization outcome depends strongly from human capital, in particular, from the players? skill. The football labor
market offers a peculiar opportunity to study incentives and human capital accumulation. Clubs hiring star players
usually also have a pay scale where top players earn disproportionally more than average players (Lucifora and
Simmons 2003). In Portugal, some players earn more than 10 times the average salary of the team. Employers use
wide pay scales to encourage workers to get better results, in the particular case of football, to win championships and
international competitions. However, it can lead the club to high losses if the sport results do not generate sufficient
income to meet the assumed responsibilities.
Given this context, this research work focuses on the effect that labour mobility between Portuguese clubs and leagues
has on wages. We want to apprehend if this particular sort of individuals, professional football players, are willing to work
(play) in companies with greater visibility (higher leagues) for less money and how much money or status within the
hierarchy of the company (team) they are available to accept in order to work for largest employers. The inverse actions
are also possible.
We obtain the information on players and clubs from Quadros de Pessoal (QP), a Portuguese longitudinal matched
employer-employee micro-data mandatory survey, annually collected by the Portuguese Ministry of Labor and Social
Solidarity in the last quarter of each year. Covering the seasons from 2002-2003 to 2009-2010 we identify more than
7000 players, which gives an average of almost 900 professional and semi-professional players each year. It consider
five types of movements: UP - player get transfer to a new club of a higher league; DOWN ? player get transfer to a new
club of a lower league; SAME ? player get transfer to a new club of the same league; PROM ? player stays in a club that
was just promoted; REL ? player stays in a club that was just relegated.
Results show that moving from an inferior league to a higher league (UP) renders an average wage increase of more
than 45%. This kind of move also shows that players are agreeable to lose positions in team hierarchy in order to play in
a higher league. On the other hand, move type DOWN shows an wage decrease around 45% and that there are a large
percentage of players (35%) who accept to earn less money in order to have a higher rank in the team?s hierarchy. For
those players who stays in the same league (SAME) it is expected a slightly wage increase (around 8%) and small
variations in hierarchy rank. Regarding players who stay with club after promotion (PROM) or relegation (REL), results
seem to show that there is not a premium or penalty in wage for being part of a winning or losing team. Concerning
hierarchy level, a player o stays with the club after relegation is expected to rise in the ranking.
Lazear, E., & Rosen, S. (1981). Rank-order tournaments as optimum labor contracts. The Journal of Political Economy,
89(5), 841-864.
Lucifora, C., & Simmons, R. (2003). Superstar Effects in Sport: Evidence From Italian Soccer. Journal of Sports
Economics, 4(1), 35-55.
Rosen, S. (1981). The economics of superstars. The American Economic Review, 71(5), 845?858.
Jelcodes:J60,J24
Labour mobility effect on wages: the professional football players’ case
António Sérgio Ribeiro
Centre for Management Studies of IST (CEG-IST), DEG
Av. Rovisco Pais
1050-001 Lisbon, Portugal
Phone: (+351) 964 292 660
FAX: (+351) 218 417 979
Corresponding author E-mail: [email protected]
Francisco Lima
Centre for Management Studies of IST (CEG-IST), DEG
Av. Rovisco Pais
1050-001 Lisbon, Portugal
Phone: (+351) 214 233 514
FAX: (+351) 218 417 979
E-mail: [email protected]
Labour mobility effect on wages: the professional football players’ case
Draft version for submission to the Druid Academy Conference 2013
(Please do not quote or cite without permission of the authors)
Abstract
We use detailed professional and semi-professional football players data to observe
the individual wage and pay hierarchical rank changes when football players move from one
club to another. The objective is to look over the players’ career path, from lower leagues to
Portuguese first league and the associated wage profile, before being eventually transferred to
onother country. Pay is clearly structured in ranks and football clubs spend a considerable
amount of resources, seeking an optimal pay ranking for inducing effort and further
development of individual abilities in order to achieve maximum efficiency. When changing
clubs, a player may move to a higher league, to a lower league or stay in the same league. We
also observe players who did not change clubs, but the clubs changed league tier – promoted
or relegated. The results suggest that players who move from a lower to a higher league are
willing to lose hierarchical pay positions in order to have the opportunity to play in a higher
league. Nevertheless, this does not mean that players lose money; they just lose salary rank
positions within the team. On the other hand, a player who moves from a higher league to a
lower league accept to play in a lower league earning less each month, but in exchange wins
hierarchical importance within the team.
Keywords: football, human capital, motivation, tournaments, wages
I.
Introduction
Over the last 20 years sport organizations became large companies with high revenues
regarding the growth of commercial and broadcasting revenues (Deloitte and Touche 2012).
The top 20 European clubs revenue reached more than € 4.4 billion in 2010/2011 season and
the tendency is to keep growing despite the economic problems felt in Europe. In 2010/2011
season, Portuguese club SL Benfica was the 21st club in Europe in revenue slightly passing €
100 million. Being one of the most popular sports in European countries, football clubs face
the challenge to concretize the objectives defined in the Union of European Football
Associations (UEFA) Financial Fair Play concept (UEFA 2009). Clubs managers will be
required to guarantee clubs sustainability while at the same time balance the desire of
supporters in winning trophies instead of reaching profit (Sloane 1971). The disregard of
UEFA rules can lead to the exclusion from internationals competitions organized by UEFA,
which are responsible to share € 1.34 billion in 2012/13 season, especially in UEFA
Champions league which has a significant impact on clubs revenue (Pawlowski et al. 2010).
The beginning of sports academic study goes back to the 50s with Rottenberg’s
(1956) research on baseball players’ labor market. Less than ten years later Neale (1964)
showed that the particular industry of sports boosts profitability with higher competitive
balance. This result was the reverse of other industries were highest competition minimize
profits. Some decades after, Rosen (1981) returned to the professional sport labor market
with a study on the inequality of earnings distribution among players. Almost 30 years after
Rosen, Zimbalist (2010) showed that, contrary to what may be expected, leagues without
salary cap distribute less revenue than leagues with salary cap. Unlike most of North
American professional leagues in Europe there is no salary cap.
One way to set workers salary is by tournaments. Lazear and Rosen (1981) show that
companies cannot base salary just on workers effort choices. The company must establish a
tournament between workers and determines prizes to winners and losers, independently of
the difference between outputs. This prize spread, or pay scales, works as an effort incentive,
because the winner is the worker who achieves the highest output, even if it is not possible to
know if the output has been reached due to luck, to effort, or both.
The tournament theory has been studied in several industries considering firms’ top
levels (Eriksson 1999) and in sports (Kahn 2000). There are a few sports example like golf
tournaments (Ehrenberg and Bognanno 1990), Italian football superstars (Lucifora and
Simmons 2003) and also tennis (Sunde 2009).
Since last decade football literature grew as football began to move large sums of
money and the several examples of historical clubs that faced bankruptcy after years of
losses, some were rescued by multi-millionaire investors and others simply felt into
forgetfulness in lower leagues. We can find examples of football studies relating to economic
situation and revenue (Kesenne 2007), players’ wages (Kuethe and Motamed 2010), club
performances (Haugen and Hervik 2002), clubs efficiency (Ribeiro and Lima 2011), players’
labor market (Frick 2007) and the revenue impact of clubs promotions and relegations (Noll
2002).
Sport industry is a particular example in which the organization outcome depends
strongly from human capital, in particular, from the players’ skill. The football labor market
is a peculiar opportunity to study the topic of incentives and human capital accumulation. In
theory, this is the case where wages are largely determined by the workers’ performance,
which can be observable to the employer (Kahn 2000), in our case the players’ performance.
This study proposes to use data from Quadros de Pessoal (QP), a longitudinal
matched employer-employee data set with information on Portuguese workers and firms,
collected yearly by the Ministry of Social Security. The QP information is matched with club
statistics collected between 2002/2003 and the 2009/10 seasons. During this period, we
identify almost 7000 players, an average of more than 800 players each year. The objective is
to look over the players’ career path, from lower leagues to Portuguese first league and the
associated wage profile, before being eventually transferred to other country. Pay is clearly
structured in ranks and football clubs spend a considerable amount of resources, therefore
they should seek an optimal pay ranking for inducing effort and further development of
individual abilities in order to achieve maximum efficiency.
After describing the national competition setting, the statistical information contained
in the datasets, and the econometric model, we estimate the relationships taking into account
the wage change occurring for each type of transition and use comparable control groups (as
in a difference-in-differences setting). We consider five types of movements: when a player
moves to a new club of a higher league; when a player moves to a new club of a lower
league; when the player moves to a new club in the same league; and when a player stays in
the same club that was promoted or relegated.
After the introduction and theoretical background the paper continues as follows.
Section II describes the Portuguese football organization. Section III presents the data used in
this research. In section IV we show the models tested and the results obtained. In Section V
we conclude our work.
II.
Portuguese Football Leagues
The Portuguese football league system has a hierarchical format with national and regional
leagues. All leagues are open and clubs compete for the best possible rank in order to get
promotion to a higher league and avoid relegation to the league below. The number of
promotion and relegation clubs can vary from league to league. This system allows that in
theory every club can rise from local leagues to First League, the Portuguese top league.
National leagues are divided in four tiers and the regional leagues have typically three
tiers, this value may vary depending on the number of teams in each region. We focus on the
first four levels – the national leagues. The first two are professional leagues and the
remaining two are semi-professional. In the years considered in this research the Portuguese
First and Second leagues had 18 teams, of which the last three were relegated, until 2005-2006
season. Since 2006-2007 the two top leagues have 16 teams and the last two are relegated to
the lower league. In order to reduce the number of teams in the first and second leagues in
2006, four were relegated clubs in the end of 2005-2006 first league season and only two were
promoted from the second league. In the second league there were five relegations and only
one promotion from the third league. The two semi-professional leagues are divided in several
divisions, some of them regarding clubs regional location, and in the end of the season it is
necessary to perform some promotion/relegation playoffs in order to know which clubs are
promoted and relegated. During this period, several events disrupted the normal functioning of
the promoted relegated scheme. Some clubs were relegated for financial reasons (Farense,
Salgueiros, Felgueiras, Estrela da Amadora), other gave up professional football (Alverca),
administrative irregularities (Gil Vicente) and there were clubs that were relegated due to
bribery of referees (Boavista) and for corruption convictions (Vizela). Table 1 summarizes the
number of clubs, divisions, promoted and relegated clubs during the seasons covered by our
study.
Table 1 – Professional and Semi-Professional leagues summary: number of clubs, relegations
promotions, and divisions.
First League
Second League
Third League
Fourth League
Season
Clubs
Rel. Clubs Prom. Rel. Divisions Clubs Prom. Rel. Divisions Clubs Prom. Rel.
2002-03
18
3
18
3
3
3
59
3
12
7
117
13
22
2003-04
18
3
18
3
3
3
59
3
11
7
118
13
27
2004-05
18
3
18
3
3
3
59
3
11
7
118
13
27
2005-06
18
4
18
2
5
4
58
1
13
7
115
13
23
2006-07
16
2
16
2
2
4
56
2
12
7
104
13
30
2007-08
16
2
16
2
2
4
55
2
15
7
94
13
28
2008-09
16
2
16
2
2
4
47
2
15
7
92
13
28
2009-10
16
2
16
2
2
3
47
2
13
8
94
13
25
Over the last 20 years, Portuguese clubs have been a source of high quality players to most
important European leagues.1 In the last years there were also a new trend, international
players, mostly South American, who came to Portugal play two or three years and get
transfer for European countries.2 To sum up to the quality that Portuguese players and players
from Portuguese clubs it is also necessary to note that Portuguese clubs play an important
role in the players and coaches transfer market. Portugal is nowadays an important source of
1
In the 1990s, for example, players like Fernando Couto, Luís Figo, Paulo Sousa, Rui Barros, Vítor Baía, Sérgio
Conceição and Carlos Secretário were transferred from Portuguese teams, which have won domestic and
international titles with European clubs. In the new millennium there are more examples of Portuguese
players who won domestic and international titles with foreign clubs (Nani, Cristiano Ronaldo, Rui Costa, Paulo
Ferreira, José Bosingwa, Deco, Tiago, Simão Sabrosa and Raul Meireles).
2
Once again there are several examples of international players that transfer from Portuguese clubs that won
domestic and international titles with clubs all over the Europe, as: David Luíz, Anderson, Radamel Falcão and
Ramires.
players and coaches not just to the most important leagues in Europe but also to the Middle
East, South Asia and Western European countries. Among the ten best placed countries in the
UEFA Country ranking over the years covered by the research only Portugal, France and
Netherlands have a positive trade balance with regard to player transfers in the overwhelming
majority of the years covered. Czech Republic and Belgium first leagues also have a positive
trade balance but unlike Portugal they do not appear consistently in the top 10 of UEFA
country ranking. First leagues from countries like Spain, England, Germany, Italy, Ukraine,
Russia and Turkey have almost yearly negative balances. This facts show that Portuguese
players and Portuguese clubs play a relevant role in the formation and sale of high-quality
players for the major European leagues and clubs.
III.
Data
Data on players and clubs came from Quadros de Pessoal, a longitudinal matched employeremployee micro-data, an annual mandatory survey collected by the Portuguese Ministry of
Labor and Social Solidarity in the last quarter of each year. The survey covers all private
firms with at least one worker. We are able to identify football players among an average of
more than 3 million workers each year. After the identification of football players it is
necessary to identify the correct league in which players started the season (before the winter
transfer market). Covering the seasons from 2002-2003 to 2009-2010 we identify 6785
players, an average of almost 850 professional and semi-professional players each year. Table
2 shows players’ statistics. Players are on average 25.0 years old. There are players who
become professionals at the age of 16, and some only retire when they have more than 40
years old. The majority of players (75%) do not stay more than two years in each club. As
expected the majority of players are from professional leagues. Clubs that compete in these
leagues have more players and all players are professional.
Table 2 - Descriptive statistics.
Variable
Wage (Euro)
Age (years)
Mean
Std. Dev.
5530.198
13610.830
25.077
4.635
Tenure (years)
1.236
4.463
League
1.896
0.946
Wage Quartile
2.539
1.117
Age < 18
0.011
0.102
18<= age <=21
0.253
0.435
22<= age <=24
0.220
0.414
25<= age <=27
0.221
0.415
28<= age <=29
0.114
0.318
30<= age <=32
0.116
0.321
age >=33
0.065
0.247
Quartile 1
0.235
0.424
Quartile 2
0.254
0.435
Quartile 3
0.247
0.431
Quartile 4
0.264
0.441
First League
0.443
0.497
Second League
0.279
0.448
Third League
0.219
0.413
Fourth League
0.062
0.240
Tenure = 0
0.544
0.498
Tenure = 1
0.202
0.401
Tenure > 2
0.254
0.435
Number of players
6785
Clubs from semi-professional leagues have fewer players and not all have contracts
with the club. We have more than 3000 players from First League and just over 400 from tier
4. Table 3 resumes players and leagues for each year. There are almost 90 clubs in our
sample, some of them with only few players. It was necessary to cross information on clubs
and players in order to guarantee that players were associated to the right league. When there
were doubts about the correct league players were drop from the sample.
Table 3 – Number of players for each league and year
Year
league
Total
1
2
3
4
2002
234
152
147
59
592
2003
246
245
254
67
812
2004
321
175
175
36
698
2005
380
282
195
63
920
2006
420
266
210
55
951
2007
429
286
231
41
1,029
2008
491
233
159
47
930
2009
497
253
95
50
895
Total
3,009
1,892
1,466
418
6785
More than 50% of players were new hires and less than 5% of players stayed more
than three years in the same club. Considering these specific characteristics of the market,
low tenure and age, we could follow 1345 moves from nearly 900 players in the years
covered. Table 4 shows the number of moves considered in the research. We must assume
that players who earn more, are those more productive and important to clubs, those who play
more and have more influence in teams’ result.
As moves we considered five types of changes: UP - player get transfer to a new club
of a higher league; DOWN – player get transfer to a new club of a lower league; SAME –
player get transfer to a new club of the same league; PROM – player stays in a club that was
just promoted; REL – player stays in a club that was just relegated. Table 4 summarizes the
number of changes that were identified in our sample.
Table 4 – Players changes between 2002-03 and 2009-10
Number of
changes
Observations
Type of
change
Observation
s
1
598
UP
265
2
227
DOWN
310
3
80
SAME
463
4
26
PROM
215
5 or more
3
REL
159
Total
934
Total
1412
IV.
Description of change
player get transfer to a new club of a
higher league
player get transfer to a new club of a
lower league
player get transfer to a new club of the
same league
player stays in a club that was just
promoted
player stays in a club that was just
relegated
Models and Results
We identify players that repeatedly appeared in the database so that we could follow
its path along the seasons covered. All clubs were identified and the professional category of
each player verified in order to reduce errors. Nevertheless, it is not possible to identify all
players and therefore may happen that some players belong to a certain club with whom they
have contract, in some cases pays them a monthly wage, but are loaned to other clubs, in the
same or different leagues. We order players by salary for each club and year and set wherein
quartile they are. In first quartile are those who have lowest wages and on the other side, in
the fourth quartile, are those players who are more among their peers.
We started to study the move UP. It was considered an UP when a player gets
transferred to a new club of a higher league. It is possible to identify 236 moves from a lower
league to a higher league, considering that the player changed club. Almost 50% of players
leaving a club in a lower league to a higher league where, before transfer, in the top paid
players of the club. After transfer the majority of players are in the bottom half of club pay
list, more than 53%. It is possible to see in figure 1 that the number of players who are in the
second and third quartile is almost exactly the same.
Figure 1 – Before and after cumulative wage quartile in UP change.
The major difference is the number of players who were in fourth quartile and after
transfer the number of players who were in the first quartile. We can observe a shift from the
fourth quartile to the first quartile. Until now, we were only interested in knowing the
cumulative salary quartile change of players who move UP without looking for the value of
salary. It is also possible to see that 176 of the 236 that moved UP changed quartile. In total,
the 236 players that moved UP summed 145 quartiles decreases and 117 increases. The
overall number of quartile changes is -28. Nevertheless, the average salary increased almost
800€ each month, what represents more than 44% of wage increase. From 236 players, less
than 90 saw their wages fall after the transfer and around 150 increased their wages.
Concerning salary changes with respect to quartiles variation it is evident that, on
average, players who accept to be transferred to a higher league club may well have different
expectations of salary change. A player who rises in club order pay rank, in other words a
player who moves to a top quartile, can expect a doubling and in some cases the triple of
salary. A player who moves to an immediately lower quartile can also expect a wage
increase. On the other hand, when players lose more than two quartile levels the cut in wage
can exceed 35%. Players who get transfer to a higher league however stays in the same
quartile can expect a wage increase of more than 27%.
After UP we studied REL. This change is considered when players move from a top
league to a lower league, throughout a club change. In our data there are 306 REL moves.
Almost 55% of players came from lower quartiles and after transfer almost 60% of them take
up the two top quartiles. Less than 16% of those players go to first quartile, when before
move they were more than 24%. On top paid players quartile it is possible to see that the
number of players in the last quartile almost doubled, rose from 58 to 104. The total sum of
quartile gains and losses is 114; this is an evident corroboration that in REL there is a transfer
of players from lower wage quartiles to higher quartiles.
Figure 2 – Before and after cumulative wage quartile in DOWN change.
Unlike UP, in REL players on average lose 47% of salary which represents more than
1450€ each month. This loss is supported by players who accept to play in a lower league and
also accept to be in lower quartiles than they were. On average, players who lose positions in
quartile rank, and even stay in the same, lose more than 2000€ each month. Players who grow
two or more positions on quartile ranking, 70 players out of 306, can expect a wage increase.
Next we analyzed the move SAME. There are 449 players that change club in the
same league. On average players increase salary 7% which represents more than 380€ each
month. As regards to wage quartile changes, we do not observe major variations but the
tendency is to rise in hierarchy ranking of salary. Players who stay in the same quartile or
increase internal rank are expected to earn a wage increase. The increase can vary almost
400%, players who move from 1st quartile to 4th quartile, or about 52% if they climb just one
quartile.
Figure 3 – Before and after cumulative wage quartile in SAME change.
Players who decrease in wage rank can expect a wage decrease higher than 50%. This
suggests that players who perform better are willing to play in the same league for better
salaries and hierarchical progress, on the other hand there are players who accept to earn less
and lost importance within the club so they can remain playing in the same league.
The next changes are PROM and REL. In these two cases players stay from one
season to another with a promoted or relegated club. On both cases, there is an average
decrease of salary around 8%. In PROM cases, it represents to an average decrease of 170€
and in REL near 230€ each month. There are 195 players who followed the promoted club
and 159 the relegated club. Regarding the wage quartile changes, there are not significant
changes in PROM, as shown in figure 4, just a few losses in quartiles. This can be due to the
fact that promoted clubs hire new players to face competitiveness of higher leagues and to
captivate those new players they have to tempt them with higher wages.
Figure 4 – Before and after cumulative wage quartile in PROM change.
The REL case is different, players can expect an increase in wage quartile since that
players had contracts for more than one season and the club comply with the obligations set
in the past. An additional cause to this wage quartile increase is that relegated clubs can count
with less revenue and so they must consider this breaking in new hires, as we should expect
less costly than in late season.
Figure 5 – Before and after cumulative wage quartile in REL change.
On the subject of wage variation related to wage quartile variation we saw that if a
player stays in the same club after promotion (PROM) and do not lose wage quartile positions
it is expected on average an wage increase, which can be more than 60% if the wage quartile
variation is equal or higher than two levels. In contrast, if players lose wage quartile positions
it is expected a wage decrease higher than 30%.
Comparable tendency can be found with REL changes. Players who climb in wage
quartile ranking can expect a wage increase and players who fall in wage quartile ranking
expect a wage decrease. The wage variations magnitudes are similar to PROM changes. The
difference between PROM and REL is that players who stays with relegated clubs and do not
change wage quartile can earn on average 10% less comparing to an 6% increase of those
who stay with promoted clubs and maintain the same wage quartile rank. The summary of
results is available in table 5.
Table 5 – Summary of quartile variations.
UP
DOWN
SAME
PROM
REL
STAY
-55
142
59
-16
59
176
62.3%
35.9%
50.6%
52.3%
40.3%
57.3%
Quartile
Variation
37.7%
Number
of
players
64.1%
Number
of
players
49.4%
Number
of
players
47.7%
Number
of
players
59.7%
Number
of
players
42.7%
Number
of
players
-3
3.8%
1.6%
2.9%
2.1%
1.9%
2.0%
-2
11.9%
6.9%
7.8%
6.7%
2.5%
5.0%
-1
26.3%
18.0%
19.4%
21.5%
10.7%
14.1%
0
25.4%
30.1%
31.2%
43.6%
44.7%
49.7%
1
19.5%
20.6%
20.0%
14.4%
25.2%
21.9%
2
9.3%
15.4%
15.1%
9.2%
10.7%
5.5%
3
3.8%
7.5%
3.6%
2.6%
4.4%
1.8%
Sum of
quartile
changes
Salary
increase
Salary
decrease
After the five possible changes, we studied players that stay (stay) in the same club
and in the same league from one year to another. In 2175 observations of stay in only 906
players lose salary. It is possible to see that there are no significant variations in salary
quartile rankings for players who remain in the same club. Figure 6 shows this trend. The
sum of all players hierarchy gains and losses is 176, which indicates that players have
tendency to benefit from being in the same club one year to another. Regarding salary
change, players who climb in hierarchy ranking can expect salaries increases of more than
50%.
Figure 6 – Before and after cumulative wage quartile in Stay change.
In order to test the relationship between the players’ position within the pay rank and
the move to another club, we estimate a probit model for the change of club. The probability
is defined as a function of player’s wage quartile before the move, in addition to the club’s
league and player’s age. The estimation results show that players in lower wage quartiles,
namely at the second, hold a higher probability of club change as compared players in other
wage quartiles (Table 6).
Table 6. Probability of change club considering the wage quartile – Probit (marginal effects)
Variable
Number of observations
.
0.190***
(0.064)
0.168***
(0.065)
0.125*
(0.065)
3654
Chi-squared
10.17
2ndWage Quartile before change
3rd Wage Quartile before change
4th Wage Quartile before change
Standard errors between brackets. * p<0.10, ** p<0.05, *** p<0.01
We examine the initial position of the player within the pay scale at the moment of
hiring to a new club by estimating the probability of entry into a particular wage quartile
using an ordered probit model. Figure 7 show the probability of entry in top and inferior
wage quartile for each league by player’s age.
Figure 7 – Probability of entry in top and inferior wage quartile.
Results show that as players get older, and consequently more become experienced,
the probability of being hired to the top of the pay scale – the top of the wage hierarchy –
increases. The effect seems more pronounced in Portuguese first league. The wage-age shape
indicates that experience while playing football has an effect on wage rank level. In contrast,
more young players are almost intended to enter clubs in the bottom of wage rank.
It was also used an ordered probit model to test the probability achieving to the top
wage ranks if players stay in the same club for several years. Figure 8 suggest that players
have a propensity to achieve higher salaries, if they make the decision of stay in the club. In
this case we are dealing with less risky players.
Figure 8 – Probability of higher/inferior salary rank by tenure.
The next step in our analysis is to model the wage variation associated with the five
career events defined previously: up; down; same; rel and prom. We define a mincerian wage
function with fixed effects for clubs and players. The model is estimated at the first difference
for wages allowing for the characterization of the wage profile upon the players’ movements
between clubs and leagues (Table 7). The results show that lower leagues (after the player’s
move) hold a negative premium when compared with the first league. The change of club
belonging to a higher league is associated with an average increase of 0.269 log points
whereas a change to a club in a lower league is associated in an average decrease of 0.242 log
points. The comparison group comprises those players who stay in the same club in two
consecutive seasons and the club remains in the same league.
Table 7. Players’ wages - first-difference estimation
Variable
Model 1
Model 2
2nd league
-0.170***
(0.021)
-0.134***
(0.019)
-0.092***
(0.030)
-0.152***
(0.021)
-0.094***
(0.019)
-0.051*
(0.030)
0.269***
(0.060)
-0.242***
3rd-league
4th league
up
down
same
rel
prom
Constant
Number of observations
Adjusted-R2
F
0.121***
(0.012)
5052
0.02
29.3
(0.051)
-0.012
(0.045)
0.031
(0.031)
-0.092
(0.069)
0.111***
(0.011)
5052
0.04
14.8
Standard errors between brackets. * p<0.10, ** p<0.05, *** p<0.01
V.
Conclusions
Our study uses detailed data on professional football players’ wages to determine in
each quartile players are in each club for all years covered. We examine the changes between
clubs of professional football players in Portugal for eight seasons. After defining five
different possible changes, we can conclude that professional football players are willing to
lose within team importance in exchange for playing in a higher league. Nevertheless, this
hierarchy loss does not mean lower wages. The opposite is also true, players are accepting to
lose money and visibility for playing in a lower league in exchange of more importance
within the new team. We also saw that typically players who move from clubs of the same
league expect to earn a wage increase and also a rise the internal rank when we talk about
monthly salary. There are two more moves considered; players who stay in the same club
after promotion cannot expect to receive a salary bonus or a rise in internal ranking. In
contrast, players who stay in the same club after relegation in general rise in teams’ pay
hierarchy. However, this rise does not mean that players receive a salary bonus.
A common conclusion for all changes is that a player who climbs in teams’ pay
hierarchy can expect a salary increase, these means that these sorts of players have their effort
recognized in two different ways: monthly salary increase and team hierarchical importance
grow.
We showed that players between 26-30 years old are transferred least than other aged
players. We also showed that those same players are on average the players who occupy the
higher hierarchy salary quartiles. Alternatively, more young players typically are hired to
lower quartiles. These can indicate that professional football players reach career top in this
age range.
Finally, an additional conclusion is that players who stay more years in the same club
can expect an increase of probability of reaching top wage quartile levels.
Acknowledgments
This work was partially supported by the Portuguese Foundation for Science and Technology
(PTDC/EGE-ECO/101493/2008 and PTDC/EGE-ECO/118070/2010). We are also thankful to
the Portuguese Ministry of Labor and Social Solidarity for giving us access to the data used in
this research.
References
Deloitte & Touche (2012). The untouchables, Football Money League. Available at
http://www.deloitte.com/assets/DcomUnitedKingdom/Local%20Assets/Documents/Industries/Sports%20Business%20Group
/uk-sbg-dfml-2012-final.pdf (accessed 13 December 2012).
Ehrenberg, R., & Bognanno, M. (1990). The Incentive Effects of Tournaments Revisited:
Evidence from the European PGA Tour. Industrial and Labor Relations Review, 43(3),
74-88.
Eriksson, T. (1999). Executive compensation and tournament theory: empirical tests on
Danish data. Journal of Labor Economics, 17, 262–80.
Frick, B. (2007). The Football Players Labor Market: Empirical Evidence From the Major
European Leagues. Scottish Journal of Political Economy, 54(3), 422-446.
Haugen, K.K., A. Hervik. (2002). Estimating the value of the Premier League or the world’s
most profitable investment project. Applied Economics Letters, 9, 117-120.
Kahn, L. M. (2000). The Sports Business as a Labor Market Laboratory. Journal of
Economic Perspectives, 14(3), 75-94.
Kesenne, S. (2007). The peculiar international economics of professional football in Europe.
Scottish journal of political economy, 54(3), 388-400.
Kuethe, T. H., & Motamed, M. (2010). Returns to Stardom: Evidence From U.S. Major
League Soccer. Journal of Sports Economics, 11(5), 567-579.
Lazear, E., & Rosen, S. (1981). Rank-order tournaments as optimum labor contracts. The
Journal of Political Economy, 89(5), 841-864.
Lucifora, C., & Simmons, R. (2003). Superstar Effects in Sport: Evidence From Italian
Soccer. Journal of Sports Economics, 4(1), 35-55.
Neale, W. (1964). The Peculiar Economics of Professional Sports. Quarterly Journal of
Economics, 78 (1), 1-14.
Noll, R. G. (2002). The Economics of Promotion and Relegation in Sports Leagues: The Case
of English Football. Journal of Sports Economics, 3(2), 169-203.
Pawlowski, T., Breuer, C., & Hovemann, A. (2010). Top Clubs Performance and the
Competitive Situation in European Domestic Football Competitions. Journal of Sports
Economics, 11(2), 186-202.
Ribeiro, A.S. & Lima, F. (2011). Portuguese Football league efficiency and players’ wages.
Applied Economics Letters, 19(6), 599-602
Rosen, S. (1981). The economics of superstars. The American Economic Review, 71(5), 845–
858.
Rottenberg, S. (1956). The baseball Player’s Labor Market. Journal of Political Economy, 64
(3), 242-258.
Sloane, P. (1971). The economics of professional football: the football club as a utility
maximiser. Scottish Journal of Political Economy, 17, 121-146.
Sunde, U. (2009). Heterogeneity and Performance in Tournaments: A Test for Incentive
Effects using Professional Tennis Data. Applied Economics, 41(25), 3199-3208.
UEFA (2009). UEFA Executive Committee Nyon decisions. 101. Available at
http://www.uefa.com/MultimediaFiles/Download/PressRelease/uefa/UEFAMedia/88/6
7/81/886781_DOWNLOAD.pdf (accessed 13 December 2012).
Zimbalist, A. (2010). Reflections on Salary Shares and Salary Caps. Journal of Sports
Economics, 11(1), 17-28.